Data



The main statistics of the dataset are presented in Table 1.

 
Table 1. Multilingual HTR dataset statistics.

 

To access the data, you must first fill out the registration form, and you will receive a link and password to download them.

Evaluation Protocol

Evaluation will be performed using WER, which measures the ratio of errors (insertions, substitutions and deletions) in a transcript to the total words of the ground-truth. Systems will be ranked based on this metric. Ties between systems with exactly the same WER result will be broken based on the character error rate (CER) and then the time of submission of the results.

Github for the evaluation tool here: Evaluation tool

Results Submission

A submission form will be provided according to the competition schedule.